Will You Make it to Your 50th Wedding Anniversary?

My wife and I are taking the next several days off to celebrate our 50th wedding anniversary. Naturally, this has caused me to think about the probability of such an event happening. Two things of demographic importance have to occur: (1) you and your spouse both have to live that long, and (2) you have to stay married. What are the probabilities associated with these things? It turns out that the answer is not as easy to figure out as it might seem.

We were high school sweethearts and that transitioned into being college sweethearts, and we married at the end of our junior year in college at Berkeley. Technically, we were both 20 at the time, but we both turned 21 that summer of 1965. Life tables for that year (or its closest approximation) suggest that a 21 year old white male had a 56% chance of surviving another 50 years, whereas a white woman aged 21 had a 75% chance of living that many additional years. The probability then that my wife and I would both still be alive 50 years later was .56*.75=.42. However, that calculation ignores the fact that life expectancy has been improving over time. In 1965, life expectancy at birth for white males was 67.5, whereas by 2010 it had improved to 76.5. So, each year that one stays alive in such a regime, your probability of dying goes down a bit. At current levels, a 21 year old white male has a 75% chance of still being alive in 50 years, whereas a 21 year old woman has an 84% chance of survival. Together, they produce a probability of .63 that the two will jointly survive. 

Now, in terms of divorce, it turns out that those probabilities have also changed in the 50 years that my wife and I have been married, but not for the better. My wife and I married just before the divorce rate started to rise. The 1970s and 80s saw a big rise in divorce although it has since leveled off, or even dropped a bit. Currently, our best source of data about the duration of marriages in the U.S. is the U.S. Census Bureau's Survey of Income and Program Participation (SIPP), as I have noted previously. Divorce is not quite as common as many people think, but even though it is not true that 50 percent of marriages will end in divorce, the real figure appears to be not a lot lower than that. The SIPP includes a marital history and from this we can see how long respondents have been married, with the caveat that data refer to survivors, not to all people. Thus, data from the 2008 panel (the most recent data I could find), show that among men married in 1965-1969, 58 percent had made it to their 35th anniversary (the highest anniversary possible at the time for men married in those years), while only 52 percent of women had made it to the 35th. The difference is obviously due to the higher mortality among men. The data suggest, though, that roughly 50 percent of couples will not divorce in their lifetimes.

Overall, then, we can make a few back of the envelope calculations to suggest that if two people aged 21 were to marry today, there is at least a 32 percent chance that they will both survive to celebrate their 50th wedding anniversary (.75*.84*.50). If, on the other hand, you both wait until age 31 to marry, the odds drop to 17 percent (.52*.65*.50), assuming no change in life expectancy over the next 50 years.

Although it is not easy to nail down the odds of making it to your 50th wedding anniversary, my wish for you is that you and your spouse (if you have one, or when you have one) both live a long life and that you have a long (and happy) marriage.